The recent, rapid expansion of available nucleic acid sequence information has necessitated the development of methods for identifying the function of nucleic acid sequences, particularly transcribed nucleic acid sequences such as expressed sequence tags, with unknown function, in an efficient and labor-cost effective way.
To identify or validate the role of sequenced nucleic acids from plants of unknown or hypothetical function it is necessary to produce or identify plants in which those nucleic acids are either structurally or functionally inactivated. Plants wherein predetermined nucleic acid sequences are structurally inactivated can be generated using recombination technologies such as homologous recombination as described by Kempin et al. (1997). Alternatively, plants with a mutation in a predetermined nucleotide sequence can be identified by screening a saturated mutant library, such as but not limited to a T-DNA insertion library or a transposon insertion library (see e.g. Pereira and Aerts, 1998). These methodologies all require the generation of a large number of permanently altered plants, and thus are less amenable for application in high throughput methods. Moreover, the recovery of plants with recessive mutations in essential genes requires time-consuming breeding to maintain the plants in heterozygous state. Maintenance of dominant lethal mutations in essential genes is virtually impossible.
Plants with functionally inactivated predetermined nucleotide sequences can be generated in a straightforward way using methodologies wherein inhibitory RNA is generated, such as antisense or sense RNA.
The use of inhibitory RNA to reduce or abolish gene expression, also known as gene silencing, is well established in the art and is the subject of several reviews (e.g. Baulcombe 1996, Stam et al. 1997, Depicker and Van Montagu, 1997). Several patent applications relate to the practical exploitation of gene silencing.
U.S. Pat. No. 5,190,931 and EP 0 467 349 A1 describe methods and means to regulate or inhibit gene expression in a cell. These comprise incorporating into or associating with the genetic material of the cell a non-native nucleic acid sequence which is transcribed to produce an mRNA which is complementary to and capable of binding to the mRNA produced by the genetic material of that cell.
EP 0 240 208 describes a method to regulate expression of genes encoded for in plant cell genomes. This method is achieved by integration of a gene under the transcriptional control of a promoter which is functional in the host and in which the transcribed strand of DNA is complementary to the strand of DNA that is transcribed from the endogenous gene(s) one wishes to regulate.
EP 0 223 399 A1 describes methods to effect useful somatic changes in plants by causing the transcription in the plant cells of negative RNA strands which are substantially complementary to a target RNA strand. The target RNA strand can be a mRNA transcript created in gene expression, a viral RNA, or other RNA present in the plant cells. The negative RNA strand is complementary to at least a portion of the target RNA strand to inhibit its activity in vivo.
EP 0 647 715 A1 and U.S. Pat. Nos. 5,034,323, 5,231,020 and 5,283,184 describe methods and means for producing plants exhibiting desired phenotypic traits, by selecting transgenes that comprise a DNA segment operably linked to a promoter, wherein transcription products of the segment are substantially homologous to corresponding transcripts of endogenous genes, particularly endogenous flavonoid biosynthetic pathway genes.
W0 93/23551 describes methods and means for the inhibition of two or more target genes. These methods comprise introducing into the plant a single control gene, which has distinct DNA regions homologous to each of the target genes, and a promoter operative in plants adapted to transcribe from such distinct regions RNA that inhibits expression of each of the target genes.
A major disadvantage of these technologies, which hampers the exploitation thereof in high throughput gene function discovery methods, is the intrinsic unpredictability and low occurrence of the gene silencing phenomenon.
Recently, Waterhouse et al. (1998) and WO99/53050 have described methods and means to make gene silencing in plants more efficient and predictable, by simultaneous expression of both sense and antisense constructs in cells of one plant. The sense and antisense nucleic acids may be in the same transcriptional unit, so that a single RNA transcript that has self-complementarity is generated upon transcription.
Hamilton et al. (1998) describe improved silencing e.g. of tomato ACC-oxidase gene expression using a sense RNA containing two additional upstream inverted copies of its 5′ untranslated region.
WO 98/53083 describes constructs and methods for enhancing the inhibition of a target gene within an organism, involving the insertion into the gene silencing vector of an inverted repeat of all or part of a polynucleotide region within the vector.
WO01/12824 describes a method and means of reducing the expression of a target nucleic acid, and thereby altering the phenotype of an organism, especially a plant, by providing aberrant, target specific RNA to the nucleus of the host cell. The method involves chimeric constructs comprising a promoter, a DNA region encoding the target RNA, a self-splicing ribozyme and a 3′end.
It should be clear however, that the use of inhibitory RNA as a tool in reversed genetics analysis of gene function via high throughput methods, whereby the inhibitory RNA is generated from gene-silencing constructs which are stably integrated in the genome of transgenic plants, suffers from the same drawbacks as mentioned herein above.
EP 0 194 809 and U.S. Pat. No. 5,500,360 suggest the use of viral RNA vectors to produce regulatory RNA such as anti-sense RNA.
Initial exploration of the use of viral vectors to deliver inhibitory RNA into cells of plants has been described by Chapman (1991). In this publication, gene silencing constructs comprising nucleotide sequences complementary to the translated region of the GUS gene on a PVX derived viral vector were described. The experiments however, remained inconclusive as to whether gene silencing could be provoked using viral vectors for the production of inhibitory RNA.
WO 93/03161 is directed toward recombinant plant viral nucleic acids and to hosts infected thereby. The non-native nucleic acid sequence which is transcribed may be transcribed as an RNA which is capable of regulating the expression of a phenotypic trait by an anti-sense mechanism.
English et al., 1996 describe the suppression of the accumulation of a viral vector comprising a foreign nucleotide sequence in transgenic plants exhibiting silencing of nuclear genes comprising the same foreign nucleotide sequences, thus linking gene silencing and viral vectors, albeit in a reverse way as envisioned here.
Kumagai et al. (1995) described the inhibition of phytoene desaturase gene by viral delivery of antisense and sense RNA.
WO 95/34668 suggests the use of genetic constructs based on RNA viruses which replicate in the cytoplasm of cells to provide inhibitory RNA, either antisense or co-suppressor (sense) RNA.
Baulcombe et al. (1998) and Ruiz et al. (1998) describe virus-induced gene silencing of the endogenous phytoene desaturase gene (PDS) or of a green fluorescent protein transgene (GFP) in plants, using potato virus X derived vectors carrying inserts homologous to PDS and GFP, respectively. The authors further suggested that virus-induced gene silencing may develop into a novel assay of gene function, by introducing a fragment of the genome into a viral vector and inferring the function of the gene from the symptoms of the infected plants exhibiting gene silencing.
WO90/12107 describes the use of STMV vectors for the making of desired proteins in plant cells by replacing the STMV coat protein gene with a heterologous nucleotide sequence, but the use of satellite viruses for introduction of inhibitory RNA is not described.
WO0063397 describes a method for virus-induced gene silencing of endogenous genes or transgenes that makes use of a two component system made up of a satellite RNA virus vector and its corresponding helper virus.
The prior art methods based on plant viral RNA vectors described above still suffer from the following drawbacks:                the phenotype induced by the inhibitory RNA may often be weak or very weak        the phenotypic pattern induced by the inhibitory RNA may not be uniform throughout the plant tissue and may be present in a patchy or mosaic pattern        the phenotype induced by the inhibitory RNA may often be masked by symptoms caused by the virus        interpretation of the gene silencing phenotype may be hampered by intra- and inter-experimental variation        in known gene-silencing systems the silencing of genes expressed in meristematic tissues was often not possible, due to an inability of the vector or signal to enter meristems        in known gene-silencing systems, with the exception of WO0063397, the chimeric viral vectors carrying gene-silencing sequences often were themselves targets of silencing, compromising the effectiveness.        
These and other problems have been solved as described hereinafter, including the different embodiments described in the claims.